Speaker at Google Internal Conferences:

BI Tech Talk 2021: Delivered a presentation on achieving 10X+ performance improvement for PetaBytes of Android Telemetry data.

DataVerse 2022: Customer lightning talk on low-latency Android Telemetry Dogfood Sessions Pipeline.

Performance Summit 2023: Shared insights on stress testing a streaming flume pipeline for scalability and reliability.

ServerCon 2023: Presented on production hygiene for streaming flume pipelines with dual homing.

BI Tech Talks 2024: Discussed emerging trends in the data warehouse industry.

Performance and Optimization Summit 2024: Demonstrated optimized petabytes-scale data processing, accelerating telemetry queries by 10-100X and reducing costs by 10-1000X through innovative strategies and cutting-edge techniques.

Recognized for Impactful Contributions: Received two Google Perfy awards in 2024:

Gold: For eliminating duplicate Gaia logging, saving ~20 SWEs worth of resources and improving privacy

Silver: For significantly improving query performance and reducing costs through advanced data partitioning for petabytes scale data.

Focusing on Machine Learning Infrastructure for Netflix Recommendations. As part of the team, my primary focus is to build out big data analytics tools like Marathon, Spark and Hadoop on mesos platform in the AWS cloud along with supporting Jupyter and Zeppelin notebooks. These tools are at the heart of ML pipelines and analytics at Netflix and we are continuously pushing their performance and scale.

Speaker @MesosCon2018: Multi-tenant Spark workflows in Auto Scalable Mesos clusters

R&D and Engineering of various Big Data tools and technologies.

As a Hadoop security architect made the Hadoop platform PCI complaint by enabling Kerberos Authentication, Sentry policies for HDFS, Hive and Hbase, end-to-end encryption using Cloudera Navigator. Enabled multi-tenancy within the Hadoop platform using Yarn Qs and ACLs. Enabled DataStax Cassandra authentication and node-to-node and client-to-node encryption. Cassandra schema design. Hbase schema design and Hbase tuning to support peta bytes of historical data load to support faster queries with less space consumption on HDFS. Hbase bulk upload to load historical data. Disaster recovery architecture for Hadoop platform. Capacity planning for Hadoop cluster. Hadoop infrastructure architecture for Development, QA and Production environments. Spark jobs performance tuning. Restful web service development using Spring boot. F5 VIP configuration to enable load balancing and SSL offloading. Hadoop platform and application automation using Chef. Hadoop and Spark application(s) design and development. Enabled Kafka security.

Design and development of WebServices, SOA governance, Platform Architecture, Reference Architecture for Digital Platform, Infrastructure Architecture, Big Data, R&D and Engineering